Kaggle project.
Project Title: Predicting a Pulsar Star
Description: Pulsars are rare neutron stars that emit detectable radio waves, offering significant insights into space-time, the interstellar medium, and states of matter. Leveraging their unique properties, including dense composition and precise rotational periods, we can utilize them to explore various astrophysical phenomena.
In my project, I used supervised machine learning algorithms to predict whether a star is a pulsar. This comprehensive analysis included:
- Data Analysis: Initial exploration and preprocessing of data.
- Logistic Regression: Building a binary classifier.
- K-Nearest Neighbour (KNN) Classification: Implementing a distance-based approach.
- Support Vector Machine (SVM) Classification: Utilizing hyperplanes for classification.
- Naive Bayes Classification: Applying probabilistic models.
- Decision Tree Classification: Creating tree-based models for decision making.
- Random Forest Classification: Enhancing decision trees with ensemble methods.
- Model Evaluation: Assessing performance metrics to ensure accuracy and reliability.
This project provided valuable experience in applying various machine learning techniques to solve problems in the realm of astrophysics.